0
$\begingroup$

I have two different audio short samples of an instrument playing. If one of the samples has a reverb effect applied, how easy is it to detect which sample has reverb? Is this possible?

$\endgroup$
2
$\begingroup$

A reverb effect usually includes an added, delayed version of the original signal.

Calculate an autocorrelation. If there is significant amount of power at delays larger than what you'd expect for a normal-sized room (say, 1/5 of a second), that'd be a good indication.

$\endgroup$
  • 1
    $\begingroup$ reverb is different than echo but with reverb there should be higher peaks of autocorrelation at large lags then there would be without reverb. but if the reverb is subtle, i think this can be missed. other than listening tests, i am not sure of a good automated way to tell. $\endgroup$ – robert bristow-johnson Jun 11 '18 at 4:34
  • 1
    $\begingroup$ I second RBW's note. If the recording is known to have "pops" then it might be possible to do a Time Frequency analysis on that and assess the spread of the impulse caused by the reverberation. This would essentially be a mini room probing. $\endgroup$ – A_A Jun 11 '18 at 10:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.